feat: add Gibsonai query tool

crewai_tools/__init__.py

@@ -34,6 +34,7 @@
     FirecrawlCrawlWebsiteTool,
     FirecrawlScrapeWebsiteTool,
     FirecrawlSearchTool,
+    GibsonAIQueryTool,
     GithubSearchTool,
     HyperbrowserLoadTool,
     JSONSearchTool,

crewai_tools/tools/__init__.py

@@ -25,6 +25,7 @@
     FirecrawlScrapeWebsiteTool,
 )
 from .firecrawl_search_tool.firecrawl_search_tool import FirecrawlSearchTool
+from .gibsonai_query_tool.gibsonai_query_tool import GibsonAIQueryTool
 from .github_search_tool.github_search_tool import GithubSearchTool
 from .hyperbrowser_load_tool.hyperbrowser_load_tool import HyperbrowserLoadTool
 from .json_search_tool.json_search_tool import JSONSearchTool
@@ -105,6 +106,6 @@
 from .brightdata_tool import (
     BrightDataDatasetTool,
     BrightDataSearchTool,
-    BrightDataWebUnlockerTool
+    BrightDataWebUnlockerTool,
 )
 from .zapier_action_tool.zapier_action_tool import ZapierActionTools

crewai_tools/tools/gibsonai_query_tool/README.md

@@ -0,0 +1,176 @@
+# GibsonAIQueryTool
+
+## Description
+This tool enables direct SQL query execution against [GibsonAI](https://www.gibsonai.com/) serverless databases (MySQL/PostgreSQL) via the hosted Data API. The GibsonAIQueryTool provides a flexible interface for performing database operations including SELECT queries for data retrieval and INSERT/UPDATE/DELETE operations for data modification. It works with any database schema deployed in your GibsonAI project, making it a versatile tool for database interactions.
+
+## Installation
+To install the `crewai_tools` package and utilize the GibsonAIQueryTool, execute the following command in your terminal:
+
+```shell
+pip install 'crewai[tools]'
+```
+
+## Setup
+
+### 1. Set up your GibsonAI Project
+In [GibsonAI](https://app.gibsonai.com), create a new project with your desired database schema. For example:
+
+```txt
+I want to create a customer management system with the following tables:
+- "customers" table with fields (name, email, phone, created_at)
+- "orders" table with fields (customer_id, product_name, quantity, price, order_date)
+- "products" table with fields (name, description, price, stock_quantity)
+```
+
+Once generated, click `Deploy` to Production and copy the API key from the `Data API` tab.
+
+### 2. Set up Environment Variables
+Create a `.env` file in your project root and add your GibsonAI API key:
+
+```env
+GIBSONAI_API_KEY=your_project_api_key_here
+```
+
+## Example Usage
+
+```python
+from crewai_tools import GibsonAIQueryTool
+
+# Initialize the tool
+tool = GibsonAIQueryTool()
+
+# Or initialize with explicit API key
+tool = GibsonAIQueryTool(api_key="your_api_key_here")
+
+# Example queries
+# 1. Select data
+result = tool.run(query="SELECT * FROM customers WHERE email LIKE '%@gmail.com'")
+
+# 2. Insert new data
+result = tool.run(query="INSERT INTO customers (name, email, phone) VALUES ('John Doe', 'john@example.com', '555-0123')")
+
+# 3. Update existing data
+result = tool.run(query="UPDATE products SET price = 29.99 WHERE name = 'Widget'")
+
+# 4. Delete data
+result = tool.run(query="DELETE FROM orders WHERE order_date < '2023-01-01'")
+
+# 5. Parameterized query (recommended for security)
+result = tool.run(
+    query="SELECT * FROM customers WHERE name = ? AND email = ?",
+    parameters={"name": "John Doe", "email": "john@example.com"}
+)
+```
+
+## Arguments
+The GibsonAIQueryTool accepts the following arguments:
+
+- `query` (required): A string containing the SQL query to execute. This can be any valid SQL statement (SELECT, INSERT, UPDATE, DELETE)
+- `parameters` (optional): A dictionary containing parameters for parameterized queries. Use this for safer query execution to prevent SQL injection attacks.
+
+## Supported Operations
+
+### Data Retrieval (SELECT)
+```python
+# Simple select
+tool.run(query="SELECT * FROM customers")
+
+# Complex select with joins
+tool.run(query="""
+    SELECT c.name, c.email, COUNT(o.id) as order_count 
+    FROM customers c 
+    LEFT JOIN orders o ON c.id = o.customer_id 
+    GROUP BY c.id, c.name, c.email
+""")
+```
+
+### Data Modification (INSERT, UPDATE, DELETE)
+```python
+# Insert single record
+tool.run(query="INSERT INTO products (name, description, price) VALUES ('New Product', 'Description', 19.99)")
+
+# Update records
+tool.run(query="UPDATE customers SET phone = '555-9999' WHERE email = 'john@example.com'")
+
+# Delete records
+tool.run(query="DELETE FROM orders WHERE quantity = 0")
+```
+
+### Parameterized Queries (Recommended)
+```python
+# Safe parameter substitution
+tool.run(
+    query="SELECT * FROM customers WHERE created_at BETWEEN ? AND ?",
+    parameters={"start_date": "2023-01-01", "end_date": "2023-12-31"}
+)
+
+tool.run(
+    query="INSERT INTO customers (name, email, phone) VALUES (?, ?, ?)",
+    parameters={"name": "Jane Smith", "email": "jane@example.com", "phone": "555-0456"}
+)
+```
+
+## Response Format
+The tool returns formatted responses based on the query type:
+
+- **SELECT queries**: Returns row count and formatted data
+- **INSERT/UPDATE/DELETE**: Returns success message with affected row count
+- **Errors**: Returns detailed error messages for troubleshooting
+
+## Security Best Practices
+
+1. **Use parameterized queries** when dealing with user input to prevent SQL injection
+2. **Store your API key securely** in environment variables, not in code
+3. **Limit database permissions** in your GibsonAI project to only what's needed
+4. **Validate input data** before constructing queries
+5. **Use transactions** for multi-step operations when supported
+
+## Error Handling
+The tool provides comprehensive error handling for:
+- Invalid SQL syntax
+- Authentication failures
+- Network connectivity issues
+- Database constraint violations
+- Timeout errors
+
+## Integration with CrewAI Agents
+This tool is designed to work seamlessly with CrewAI agents for database-driven workflows:
+
+```python
+from crewai import Agent, Task, Crew
+from crewai_tools import GibsonAIQueryTool
+
+# Create database agent
+db_agent = Agent(
+    role='Database Analyst',
+    goal='Query and analyze customer data',
+    backstory='Expert in SQL and data analysis',
+    tools=[GibsonAIQueryTool()],
+    verbose=True
+)
+
+# Create analysis task
+analysis_task = Task(
+    description='Find the top 10 customers by order value',
+    agent=db_agent,
+    expected_output='List of top customers with their total order values'
+)
+
+# Execute the crew
+crew = Crew(agents=[db_agent], tasks=[analysis_task])
+result = crew.kickoff()
+```
+
+## Troubleshooting
+
+### Common Issues
+1. **"Missing GIBSONAI_API_KEY"**: Ensure your API key is set in the environment or passed to the constructor
+2. **"Unauthorized"**: Verify your API key is correct and has the necessary permissions
+3. **"Invalid query"**: Check your SQL syntax and ensure it matches your schema
+4. **"Connection failed"**: Check your internet connection and GibsonAI service status
+
+### Getting Help
+- Check your GibsonAI project's Data API documentation
+- Verify your database schema in the GibsonAI console
+- Test simple queries first before complex operations
+- Use parameterized queries for dynamic content

crewai_tools/tools/gibsonai_query_tool/__init__.py

@@ -0,0 +1,3 @@
+from .gibsonai_query_tool import GibsonAIQueryTool
+
+__all__ = ["GibsonAIQueryTool"]